Table 2. Strengths, weaknesses and appraisals of identified studies.

VRBPAC- Vaccine Related and Biological Product Advisory Committee

Author	Strengths	Weaknesses	NOS	Assessment
Pfizer FDA VRBPAC  [17]	Randomized controlled, pivotal study - Methodologically robust - 112-day follow-up follow-up	Subgroup analysis of previously infected relatively small (3% of overall cohort). - Industry-sponsored - 2x withdrawal by a participant in the placebo group.  Only 2/19 cases were included in subgroup analysis.	Selection: ★★★ Comparability: ★★ Outcome: ★★★	(8/9)
Moderna FDA VRBPAC  [18]	Randomized controlled, pivotal study -Methodologically robust - 104-day follow-up	Subgroup analysis of previously infected relatively small (0.15% of overall cohort). - Industry-sponsored	Selection: ★★★★ Comparability: ★★ Outcome: ★★★	(9/9)
J&J FDA VRBPAC  [19]	Randomized controlled, pivotal study -Methodologically robust - 125-day follow-up	Subgroup analysis of previously infected relatively small (3% of overall cohort). - Timing of reinfections not reported related to vaccination dosing. - Industry-sponsored	Selection: ★★★★ Comparability: ★★ Outcome: ★★★	(9/9)
Goldberg et al. [20]	A large whole population study conducted in Israel - Statistically robust,  adjusting for individual cohort dynamics. - Reported outcomes by age and severity of illness. - Study period during Israeli B.1.17 surge.	3 month follow-up period - Did not compare vaccination efficacy in previously infected -Limited to PZ vaccination only.	Selection: ★★★ Comparability: ★★ Outcome: ★★★	(8/9)
Shrestha et al. [21]	Large observational study of HCWs - a 5-month observational period - Specifically attempts to answer the research question of vaccine efficacy in previously infected. - Included both  PZ, Moderna vaccines. - Adjusted for a phase of an epidemic -Rigorous testing protocol	Based on HCWs in 1 large U.S.  health system - Despite a large overall sample size, the previously infected cohort was underpowered. -Studied only symptomatic infection.	Selection: ★★ Comparability: ★★ Outcome: ★★★	(7/9)
Lumley et al. [22]	High-risk exposure group (HCWs) - Serology used to confirm exposure and control groups. - Adjusted for days at risk,    demographics,   staff occupational   role and patient  contact -Considered variants of concern   (B.1.1.7), symptomatic illness	Limited to HCWs, so may not be generalizable, particularly in children or elderly populations. -Underpowered to resolve a difference between vaccinated/ seronegative and seropositive groups.	Selection: ★★ Comparability: ★★ Outcome: ★★★	(7/9)
Cavanaugh et al. [23]	Specifically attempts to answer the research question of vaccine efficacy in previously infected. - Focus on the general population	Limited to two months (May/June) in one state. - Not a test-negative design, underestimating possible infection. - Not controlled for serological status, only history of the prior test. - Not powered to elucidate subgroup trends. -  Case-control methodology disallows for calculation of absolute risk reduction.	Selection: ★ Comparability: ★★ Exposure: ★	(4/9)
Satwik et al. [24]	An observational study, performed during the Indian Delta variant surge. - Stratified results by the severity of disease, and a number of vaccine doses. - 5-month observational period	Based on HCWs in  1 tertiary Indian hospital; a small sample size. - Limited to AZ vaccine use only. - Studied only symptomatic infection -	Selection: ★★ Comparability: ★★ Outcome: ★★★	(7/9)
Gazit et al. [25]	1:1 matched cohort study with a high volume of subjects from a national database -Adjusted for age, gender, geographic area, socioeconomic status - Designed to answer questions pertinent to policy ( protection of the previous infection vs. vaccination, the durability of protection, and risk reduction of vaccination in previously infected. -  Timed during Delta surge in Israel	Retrospective in nature. - Results may specifically reflect Delta strain phenomena - Assessed only Pfizer vaccine	Selection: ★★★ Comparability: ★★ Outcome: ★★★	(8/9)